Interesting discussion btwn Shor, @gelliottmorris, @mattyglesias on why referenda & legislation fail despite policy polling showing big support (eg, on gun background checks)
I'll add as an academic: divergence btwn item polls & outcomes can be consistent with 'unbiased' polls.
I'll add as an academic: divergence btwn item polls & outcomes can be consistent with 'unbiased' polls.
https://twitter.com/davidshor/status/1383850170505302029
Public policy battles are different from polling: partisan & group cues, media, market as prison & structural power, tradeoffs, etc...
Divergence btwn pre-policy battle polls & outcomes doesn't necessarily mean "polls bad." It can teach us a lot about the policy process
Divergence btwn pre-policy battle polls & outcomes doesn't necessarily mean "polls bad." It can teach us a lot about the policy process
Shor is saying polls should do more to recreate realism of policy battles via partisan frames and such. There's value in that, but in a lot of settings we actually care about what people say when they're just drawing on considerations they already have
Last thing I'll add--it's not a strong argument to say "these liberal researchers/groups have an ideology." There's also a huge market for "guy telling hard truths to naive liberal idealists." So everybody could be doing this instrumentally, or nobody could.
Michael Tesler has shown us the difference between ACA polling and ACA-with-Black-POTUS polling. We need both polls to understand politics, and the difference between them is probably not “bias” but rather politics we need to understand
• • •
Missing some Tweet in this thread? You can try to
force a refresh
